Bubble nucleation in magmas: A dominantly heterogeneous process?

Abstract

Bubble nucleation is at the heart of magma ascent and degassing. Nucleation kinetics control how easily a magma exsolves its volatiles, and thereby its explosivity close to the surface. Homogeneous nucleation (i.e. without assistance from crystal substrates or other phases) often requires attainment of large supersaturation pressures (ΔPN>100MPa, the change in pressure required for onset of nucleation), while heterogeneous nucleation on pre-existing crystals can occur at significantly lower values (ΔPN<50MPa). Experiments have shown that both nucleation mechanisms are viable in the laboratory. The extent, however, to which one or the other nucleation mechanism dominates in nature is still unresolved. Yet, this distinction is fundamental for applications that employ nucleation theory and its derivatives. The classical nucleation theory allows calculations of nucleation rates in silicate melts, or conversely, yields valuable estimates of supersaturation pressures ΔPN given experimental constraints on pressure conditions at the onset of nucleation. Decompression-rate meters (e.g. Toramaru, 2006) also relate the number density of bubbles in pyroclasts – a readily obtained textural characteristic – to rates of pressure change (dP/dt) during ascent. All these treatments require an explicit or implicit decision as to whether magma degassing was dominated by homogeneous or heterogeneous nucleation, which is guided by the choice of surface tension values. This contribution exploits the current textural datasets available for eruptions involving a variety of magma compositions to test the assumptions that homogeneous and/or heterogeneous nucleation dominate bubble vesiculation in natural melts. The comparison between pyroclast textures and those obtained via decompression experiments in the laboratory indicates that supersaturation pressures and decompression rates required for homogeneous nucleation are unrealistically high for natural magmas. I review the arguments for and against homogenous nucleation, and defend the alternate view that heterogeneous nucleation may dominate in all magma types, including rhyolites. In order to solve this central debate, our focus needs to be shifted towards understanding the pre-decompression phase equilibria and textural characteristics of magnetite in addition to the state of the melt structure.